2019 ComplementaryRecommendationsABr

• (Yu, Litchfield et al., 2019) ⇒ Hang Yu, Lester Litchfield, Thomas Kernreiter, Seamus Jolly, and Kathryn Hempstalk. (2019). “Complementary Recommendations: A Brief Survey.” In: The Proceedings of the 2019 International Conference on High Performance Big Data and Intelligent Systems (HPBD&IS). DOI: 10.1109/HPBDIS.2019.8735479.

Subject Headings: Complementary Recommendations.

Cited By

• http://scholar.google.com/scholar?q=%222019%22+Complementary+Recommendations%3A+A+Brief+Survey

Quotes

Abstract

Driven by the rapid development of global e-commerce, recommender systems have become a hot topic across both industry and academia because they offer a potential source of business revenue. A wide range of recommendation solutions have been developed and they can be classified as substitute and complementary: substitute recommenders offer similar items to the source item, complementary recommenders suggest items that are dissimilar to the source but are often sold as a companion item or service (such as a mobile phone (source) and case (complementary)). Both types of recommendations demonstrate unique values in specific application domains. However, the research and development of complementary recommendations still remain sparse when compared with substitute recommenders. This paper presents a brief survey on existing solutions for complementary recommendations. Our work summarizes the existing research activities and explores open questions in this field by discussing three aspects including the identification of the ground truth for complements; the recommendation models; and evaluation datasets and metrics. To the best of our knowledge, this work is one of the few surveys that provide particular insights on complementary recommendations in recent years.

I. Introduction

Recommender systems are increasingly attracting the attention of both industry and academia because of the potential value generated in e-commerce applications [ 1 ]. For recommender systems, the ultimate target is to [[stimulate the purchasing activities of potential customers by retrieving the items that catch their personalized interests among the overloaded information. With this goal, such systems can be classified as substitute and complementary recommenders: substitute recommenders offer similar items to the source, complementary recommenders suggest items that are dissimilar to the source but are often sold with it as a companion item or service (such as a mobile phone (source) and case (complementary) ).

Compared with the deep and mature research outcomes on substitute recommenders, there still exist research gaps and opportunities for complementary recommenders. However, the research and development of complementary recommenders face unique challenges that require specific methodologies to address; the most typical challenge resides in complements discovery that is non-trivial for the system.

Fig. 1. Publications over year

Unlike substitutes that could be explicitly identified as similar products interchangeably viewed or purchased, a large number of complementary products are loosely bound by latent principles that are difficult to capture effortlessly. As a result, the relationships might be inappropriately defined as opposed to the ground truth, which imposes extra pressure on the model training and tuning. Correspondingly, from the perspective of modeling, the high complexity of complementary relationships may limit the direct application of some substitute recommendation models such as the ones that rely on simple similarity measures considering the possibly long distance between two complements in feature space [2]. The above challenges need to be addressed considering the high value of complementary recommendations [3]. This paper surveys 23 publications on complementary recommendations published between 1994 and 2018 including two publication sources that are conference proceedings and journals. The time distribution of all the publications, displayed in Fig. 1 and surveyed in this paper, shows that the research on complementary recommendations remains sparse before 2014 but demonstrates an increased popularity afterwards. Conference publications account for 87% of all papers surveyed.

In comparison with other surveys that mainly focus on substitute recommendations [4]–[9], we aim to contribute specific insights on complementary recommendations to future researchers in the following aspects:

1) The identification of the ground truth for complementary products.

2) Currently available complementary recommendation models.

3) Currently available evaluation data sets and metrics for complementary recommenders.

The rest of the paper is organized as follows: in Section II, the identification of the ground truth for complementary products is introduced. Next, review of modeling including unsupervised and supervised learning is presented in Section III, followed by the summary of the typical evaluation data and metrics in Section IV. Finally, the conclusions are presented in Section V.

II. Identifying the Ground Truth for complementary products

…

III. Modeling complementary relationships

…

IV. Complementary recommender Evaluation

…

V. Conclusions

Recommender systems, which can be classified as substitute and complementary types, are one of the most effective solutions to boost business revenue. Therefore, the research and development of this topic have been rapidly progressing in recent years. Compared with the majority of the research that focuses on substitute recommenders, this survey reviews the currently available solutions and provides specific insights for complementary recommendations in three aspects including identifying the ground truth for complements, modeling complementary relationships, and evaluating complementary recommenders. The information aggregated in our work could provide researchers with a solid reference that enables future advancement for this field.

References

[1] K. Hosanagar, D. Fleder, D. Lee, and A. Buja, “Will the global village fracture into tribes? recommender systems and their effects on consumer fragmentation,” Management Science, vol. 60, no. 4, pp. 805–823, 2013.

[2] T. Zhao, J. McAuley, M. Li, and I. King, “Improving recommendation accuracy using networks of substitutable and complementary products,” in 2017 International Joint Conference on Neural Networks (IJCNN), May 2017, pp. 3649–3655.

[3] M. Zhang and J. Bockstedt, “Complements and substitutes in product recommendations: The differential effects on consumers’ willingness- to-pay,” CEUR Workshop Proceedings, vol. 1679, pp. 36–43, 2016.

[4] J. Bobadilla, F. Ortega, A. Hernando, and A. Gutirrez, “Recommender systems survey,” Knowledge-Based Systems, vol. 46, pp. 109 – 132, 2013. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0950705113001044

[5] J. Beel, B. Gipp, S. Langer, and C. Breitinger, “Research-paper recommender systems: a literature survey,” International Journal on Digital Libraries, vol. 17, no. 4, pp. 305–338, Nov 2016. [Online]. Available: https://doi.org/10.1007/s00799-015-0156-0

[6] M. Kunaver and T. Pozˇrl, “Diversity in recommender systems–a survey,” Knowledge-Based Systems, vol. 123, pp. 154–162, 2017.

[7] Z. Batmaz, A. Yurekli, A. Bilge, and C. Kaleli, “A review on deep learning for recommender systems: challenges and remedies,” Artificial Intelligence Review, Aug 2018. [Online]. Available: https://doi.org/10.1007/s10462-018-9654-y

[8] J. Bentahar, M. Taghavi, K. Bakhtiyari, and C. Hanachi, “New Insights Towards Developing Recommender Systems,” The Computer Journal, vol. 61, no. 3, pp. 319–348, 06 2017.

[9] I. Portugal, P. Alencar, and D. Cowan, “The use of machine learning algorithms in recommender systems: systematic review,” Expert Systems with Applications, vol. 97, pp. 205 – 227, 2018. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0957417417308333

[10] J. Zheng, X. Wu, J. Niu, and A. Bolivar, “Substitutes or complements: Another step forward in recommendations,” in Proceedings of the 10th ACM Conference on Electronic Commerce, ser. EC ’09. New York, NY, USA: ACM, 2009, pp. 139–146. [Online]. Available: http://doi.acm.org/10.1145/1566374.1566394

[11] T. Raeder and N. V. Chawla, “Market basket analysis with networks,” Social Network Analysis and Mining, vol. 1, no. 2, pp. 97–113, Apr 2011. [Online]. Available: https://doi.org/10.1007/s13278-010-0003-7

[12] I. Trofimov, “Inferring complementary products from baskets and browsing sessions,” CoRR, vol. abs/1809.09621, 2018.

[13] A. Veit, B. Kovacs, S. Bell, J. McAuley, K. Bala, and S. Belongie, “Learning visual clothing style with heterogeneous dyadic co- occurrences,” in Proceedings of the 2015 IEEE International Conference on Computer Vision (ICCV), ser. ICCV ’15. Washington, DC, USA: IEEE Computer Society, 2015, pp. 4642–4650. [Online]. Available: http://dx.doi.org/10.1109/ICCV.2015.527

[14] Y. Zhang, H. Lu, W. Niu, and J. Caverlee, “Quality-aware neural complementary item recommendation,” in Proceedings of the 12th ACM Conference on Recommender Systems, ser. RecSys ’18. New York, NY, USA: ACM, 2018, pp. 77–85. [Online]. Available: http://doi.acm.org/10.1145/3240323.3240368

[15] J. McAuley, R. Pandey, and J. Leskovec, “Inferring networks of substitutable and complementary products,” in Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, ser. KDD ’15. New York, NY, USA: ACM, 2015, pp. 785–794. [Online]. Available: http://doi.acm.org/10.1145/2783258.2783381

[16] M. Rudolph, F. Ruiz, S. Mandt, and D. Blei, “Exponential family embeddings,” in Advances in Neural Information Processing Systems 29, D. D. Lee, M. Sugiyama, U. V. Luxburg, I. Guyon, and R. Garnett, Eds. Curran Associates, Inc., 2016, pp. 478–486. [Online]. Available: http://papers.nips.cc/paper/6571-exponential-family-embeddings.pdf

[17] J. McAuley, C. Targett, Q. Shi, and A. van den Hengel, “Image- based recommendations on styles and substitutes,” in Proceedings of the 38th International ACM SIGIR Conference on Research and Development in Information Retrieval, ser. SIGIR ’15. New York, NY, USA: ACM, 2015, pp. 43–52. [Online]. Available: http://doi.acm.org/10.1145/2766462.2767755

[18] Z. Wang, Z. Jiang, Z. Ren, J. Tang, and D. Yin, “A path-constrained framework for discriminating substitutable and complementary products in e-commerce,” in Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining, ser. WSDM ’18. New York, NY, USA: ACM, 2018, pp. 619–627. [Online]. Available: http://doi.acm.org/10.1145/3159652.3159710

[19] R. He, C. Packer, and J. McAuley, “Learning compatibility across categories for heterogeneous item recommendation,” in 2016 IEEE 16th International Conference on Data Mining (ICDM), Dec 2016, pp. 937–942.

[20] E. Smirnova and F. Vasile, “Contextual sequence modeling for recommendation with recurrent neural networks,” in Proceedings of the 2nd Workshop on Deep Learning for Recommender Systems, ser. DLRS 2017. New York, NY, USA: ACM, 2017, pp. 2–9. [Online]. Available: http://doi.acm.org/10.1145/3125486.3125488

[22] M. J. Zaki, “Generating non-redundant association rules,” in Proceedings of the Sixth ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, ser. KDD ’00. New [21] R. Agrawal and R. Srikant, “Fast algorithms for mining association rules in large databases,” in Proceedings of the 20th International Conference on Very Large Data Bases, ser. VLDB ’94. San Francisco, CA, USA: Morgan Kaufmann Publishers Inc., 1994, pp. 487–499. [Online]. Available: http://dl.acm.org/citation.cfm?id=645920.672836 York, NY, USA: ACM, 2000, pp. 34–43. [Online]. Available: http://doi.acm.org/10.1145/347090.347101

[23] K. Gouda and M. J. Zaki, “Efficiently mining maximal frequent itemsets,” in Proceedings 2001 IEEE International Conference on Data Mining, Nov 2001, pp. 163–170.

[24] P.-N. Tan, V. Kumar, and J. Srivastava, “Selecting the right objective measure for association analysis,” Information Systems, vol. 29, no. 4, pp. 293 – 313, 2004, knowledge Discovery and Data Mining (KDD 2002). [Online]. Available: http://www.sciencedirect.com/science/article/pii/S0306437903000723

[25] G. Pandey, S. Chawla, S. Poon, B. Arunasalam, and J. G. Davis, “Association rules network: Definition and applications,” Statistical Analysis and Data Mining: The ASA Data Science Journal, vol. 1, no. 4, pp. 260–279. [Online]. Available: https://onlinelibrary.wiley.com/doi/abs/10.1002/sam.10027

[26] H. Hwangbo, Y. S. Kim, and K. J. Cha, “Rec- ommendation system development for fashion retail e-commerce,” Electronic Commerce Research and Applications, vol. 28, pp. 94 – 101, 2018. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S1567422318300152

[27] J. Han, J. Pei, and Y. Yin, “Mining frequent patterns without candidate generation,” in Proceedings of the 2000 ACM SIGMOD International Conference on Management of Data, ser. SIGMOD ’00. New York, NY, USA: ACM, 2000, pp. 1–12. [Online]. Available: http://doi.acm.org/10.1145/342009.335372

[28] H. Zhao, L. Si, X. Li, and Q. Zhang, “Recommending complementary products in e-commerce push notifications with a mixture model approach,” in Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval, ser. SIGIR ’17. New York, NY, USA: ACM, 2017, pp. 909–912. [Online]. Available: http://doi.acm.org/10.1145/3077136.3080676

[29] M. Grbovic, V. Radosavljevic, N. Djuric, N. Bhamidipati, J. Savla, V. Bhagwan, and D. Sharp, “E-commerce in your inbox: Product recommendations at scale,” in Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, ser. KDD ’15. New York, NY, USA: ACM, 2015, pp. 1809–1818. [Online]. Available: http://doi.acm.org/10.1145/2783258.2788627

[30] C. P. Huynh, A. Ciptadi, A. Tyagi, and A. Agrawal, “Craft: Comple- mentary recommendation by adversarial feature transform,” in Com- puter Vision – ECCV 2018 Workshops, L. Leal-Taixe´ and S. Roth, Eds. Cham: Springer International Publishing, 2019, pp. 54–66.

[31] I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y. Bengio, “Generative adversarial nets,” in Advances in Neural Information Processing Systems 27, Z. Ghahramani, M. Welling, C. Cortes, N. D. Lawrence, and K. Q. Weinberger, Eds. Curran Associates, Inc., 2014, pp. 2672–2680. [Online]. Available: http://papers.nips.cc/paper/5423- generative-adversarial-nets.pdf

[32] K. Zhao, X. Hu, J. Bu, and C. Wang, “Deep style match for comple- mentary recommendation,” in AAAI Workshops, 2017. [Online]. Avail- able: https://aaai.org/ocs/index.php/WS/AAAIW17/paper/view/15069

[33] D. M. Blei, A. Y. Ng, and M. I. Jordan, “Latent dirichlet allocation,” Journal of machine Learning research, vol. 3, no. Jan, pp. 993–1022, 2003.

[34] O. Barkan and N. Koenigstein, “Item2vec: Neural item embedding for collaborative filtering,” in 2016 IEEE 26th International Workshop on Machine Learning for Signal Processing (MLSP), Sep. 2016, pp. 1–6.

[35] G. C. Tomas Mikolov, Kai Chen and J. Dean, “Efficient estimation of word representations in vector space,” in 1st International Conference on Learning Representations, ICLR 2013, Scottsdale, Arizona, USA, May 2-4, 2013, Workshop Track Proceedings, 2013. [Online]. Available: http://arxiv.org/abs/1301.3781

[36] X. Han, Z. Wu, Y.-G. Jiang, and L. S. Davis, “Learning fashion compatibility with bidirectional lstms,” in Proceedings of the 25th ACM International Conference on Multimedia, ser. MM ’17. New York, NY, USA: ACM, 2017, pp. 1078–1086. [Online]. Available: http://doi.acm.org/10.1145/3123266.3123394;

	Author	volume	Date Value	title	type	journal	titleUrl	doi	note	year
2019 ComplementaryRecommendationsABr	Hang Yu Lester Litchfield Thomas Kernreiter Seamus Jolly Kathryn Hempstalk			Complementary Recommendations: A Brief Survey						2019